RESUMO
Timely administration of appropriate empirical antibiotics in febrile neutropenia is crucial for favourable patient outcomes. There are guidelines in place recommending such antibiotics. However, regional variations and local epidemiological data must be evaluated to tailor the antibiotics for best possible and rational use. In this study, we audited the clinical and microbiological data of febrile neutropenic episodes occurring at a tertiary care haematology institution. Three hundred and ninety-three febrile neutropenic episodes occurring in 123 patients over a 1-year period were analysed for microbial profile, sensitivity and resistance patterns, and finally clinical outcomes. Gram-negative bacilli (GNB) blood stream infections (46.9%) were more prevalent as compared to gram-positive infections (41.9%). Overall mortality due to complicated neutropenic sepsis was 19.5% (24/123 patients). Increased resistance to carbapenems, beta-lactam beta-lactamase inhibitor combinations, aminoglycosides, fluoroquinolones, and cephalosporins were observed. Cefepime and tigecycline resistance were seen in 20% and 15% GNB isolates, respectively. Chest was the most frequent focus of infection, and acute myeloid leukaemia (AML) was the most common underlying disorder which correlated with the likelihood of death (p < 0.01). Multidrug-resistant GNB (esp. Klebsiella sp.) are still most worrisome isolates in neutropenic patients. Single-agent cefepime or piperacillin-tazobactam/tigecycline combination may be considered empirical agents. Chest infections and AML were independent predictors of poor clinical outcome in neutropenic patients. Regular audit of infections and antibiotic susceptibility data is needed to improve clinical outcomes in patients with febrile neutropenia.